The OP is right that there are other factors here. A stronger word study is needed with better tools than Strongs, and I'm afraid the other factors might not have been entirely positive. But see what you think:-
Of course a Hebrew word won't be likely to map across perfectly into the same single English word every time.
דֵּעַ didn't open questions about God's omniscience. That's arising largely from English making knowledge and opinion into contrasting things - which it does with lots of pairs of words, to make up for its general lack of syntactic markers and particles. English has to do things by context and having most of the words in the shared vocabulary implying some converse or opposite helps it to get away with that.
Strongs lexicon is pretty futile for this particular word study, because it only counts the words in the work that's trying to be studied. A list of other places where the book of Job uses dei can't tell us much.
What's needed from a lexicon is what other writers used the word for. Have cognate forms been checked? What about Aramaic and Chaldean? What does the LXX use?
Gesenius has more usages:-
https://archive.org/details/geseniushebrewch0000samu_l2q1/page/204/mode/2up?q=+%D7%93%D6%B5%D6%BC%D7%A2%D6%B7
Isaiah 11:9 - - the earth shall be full of the knowledge of the Lord as the waters cover the sea
Isaiah 28:9 - - To whom will he teach knowledge, and to whom will he explain the message?
So that's דֵּעַ being used of God's knowledge being taught to a mortal, with the emphasis being on difficulty of transmission.
Psalm 73:11 - - And they say, “How can God know? Is there knowledge in the Most High?”
Proverbs 24:14 - - Know דְּעֶ֥ה that wisdom is such to your soul; if you find it, there will be a future, and your hope will not be cut off.
Jeremiah 22:16 - - Is not this to know הַדַּ֥עַת me? declares the Lord.
So back to these places where the translations read opinion:-
Job 32:6 - and afraid to tell you what I think (dei).
Some translations read 'knowledge' and the opinion is that knowledge comes from God... so probably knowledge is quite safe!
Job 32:10 - to me, I too will tell what I think (dei).
Job 32:17 - my share, I also will tell my opinion (dei).
These two should be treated as a single usage because v17 is a run-on from v10 and by v17 he still hasn't said what he knows yet.
Again, some translations do read 'know' and what he says he knows is that (after a huge long build-up) God made him and breathed life into him. So again that's pretty safe to read 'knowledge'. It might be claimed knowledge, he might be starting a lie with a truth, but that's the word he used.
And a very good reason for translating that way is that knowledge is our language's most neutral-register term for mental inventory, whereas opinion has a diminished/disparaged register. It's a word we should be careful with.
Did the LXX agree?
Job 32:6 - - ἐπιστήμην
Job 32:10 - - ἃ οἶδα
Job 32:17 - - (not rendered in translation)
Job 36:3 - - ἐπιστήμην
Job 37:16 - - ἐπίσταται
Isaiah 11:9 - - τοῦ γνῶναι
Isaiah 28:9 - - ἀγγελίαν
Psalm 73:11 - - γνῶσις
Proverbs 24:14 - - σοφίαν
Jeremiah 22:16 - - τὸ μὴ γνῶναί
(The reason for consulting the LXX is we've a far larger corpus of surviving Greek texts, so we can be surer of what a Greek writer means by ἐπιστήμη than what a Hebrew writer means by דֵּעַ.)
The usages in Job are being rendered consistently in Greek with ἐπιστήμη which is knowledge, particularly gained by experience or acquaintance with a matter or scientific knowledge (Liddell & Scott 1897, p.555).
The related Hebrew forms attract slightly different Greek translations, but γνῶσις and σοφία are firmly in the same mental realm as ἐπιστήμη and hold the same presumed truth-value.
So it starts to look like the translations that have Elihu "thinking that" or "opining" might be inventing a contrast with the omniscience of God, perhaps because they want to look pious by putting down Job's opponent. Or wanting to soften even a momentary conflict between two righteous positions - because y'know: home groups, things are simpler in black and white, don't scare the fish, et cetera. Which would be indefensible translation practice - it looks to be inherited from KJV and even makes its way into ESV and YLT. I am agog.
The contrast is surely there - but it's brought out by Job's using the same word not a different one.
The commentaries tend to buttress this criticism of the translations, with Gills, Jamieson-Fausset-Brown, and Keil and Delitzsch qualifying "opinion" as "knowledge" and the Pulpit Commentary as "conviction".
UPDATE (19/08/24 - 21:55)
LXX have λαλήσω in two places, so I suppose they might have found a smidgen of negative register on one of the other Hebrew words, and shifted it to this verb of speech, and then KJV do something similar. That would need a wider study of the passage I'll have a look and edit if I can find anything. 32:17 חֶלְקִ֑י can imply 'seductive' speech but the derived sense of a portion of the speaking time seems natural enough to discount that.